The Last Problem AI Will Solve Is Health

Technology is a force that creates abundance, and as it rapidly solves problems in software and intelligence, the now scarcer domain of health and longevity is emerging as the next frontier.

Many industries have already approached an asymptote, where the perceived gains from further improvement are shrinking. Like TVs, Uber, smartphones, and AI models, performance keeps improving, but users can no longer feel much of a difference. Bio, however, is not yet at that optimization stage — it remains at the stage of solving fundamental problems.

Examples of bio innovation cited include:

• Monoclonal antibodies: starting out as a lab tool, they grew into an industrial platform, and now over 100 therapies have been approved for cancer, autoimmune, cardiovascular, and infectious diseases.
• Genome sequencing cost: it has fallen a millionfold over the past 25 years, enabling companies and treatment approaches that were previously impossible.
• Cell therapy and gene therapy: for some cancers, these can even offer potential cures, but the cost per patient is still too high and scalability remains limited.

Health is the most important form of abundance, and in particular, healthy time itself is the core asset. Even in developed countries, many people spend the last 20 years of their lives dealing with disease and the complications of aging, and current healthcare systems are still more skewed toward managing disease than toward prevention and maintenance.

This piece argues that AI can lower the cost of bio through understanding biology, analyzing data, reducing risk, and accelerating discovery. If that happens, highly skilled medical care, early diagnosis, and curative treatments will increasingly open up to a broader range of people, and talent, capital, and innovation are likely to concentrate on bio/health.